Axial Attention Inside a U-Net for Semantic Segmentation of 3D Sparse LiDAR Point Clouds.
Tang-Kai YinLiang-Yue WuTzung-Pei HongPublished in: IV (2022)
Keyphrases
- point cloud
- semantic segmentation
- superpixels
- structure from motion
- weakly supervised
- surface reconstruction
- conditional random fields
- scene classification
- point sets
- lidar data
- point cloud data
- object classes
- object categories
- high dimensional
- multiscale
- long range
- viewpoint
- object recognition
- image set
- visual attention